A Dynamic Incremental Rule Extracting Algorithm Based on the Improved Discernibility Matrix

Yong Liu, Xuelan Li, Congfu Xu, Yunhe Pan

Abstract: Although the rough set theory is a kind of very useful mathematical tools to deal with vagueness, uncertainty, and impression information, it is relatively difficult to be applied to the analysis of incremental data sets. In this paper, a rule extracting algorithm from incremental data sets based on the improved discernibility matrix is proposed. We introduce the belief measure to the rule sets of this algorithm is that it does not need to re-compute overall data with increment. Finally, we present an example to illustrate the main characteristics of this new incremental algorithm.